Formulaic Imaging for Tissue Diagnosis

Case ID:

This patented system (U.S. Patent 9,952,157) uses a combination of green and blue fluorescence with reflectance, aimed at simple but effective contrast improvements through division, multiplication, subtraction and additions of individual images. To the best of our knowledge no other formulaic images system based on three or more components uses near-UV or mid-UV excitation.

Observation is conducted at one or multiple wavelength bands ranging from 300-800 nm. Signal is related to emission of native fluorophores such as tryptophan, NADH, collagen, elastin and FAD and light that is reflected and/or has undergone multiple scattering and absorption events on native chromophores and scatterers. The formulaic computations are conducted based on a single, two and more images recorded at different optical configuration and named here A, B, C for illustration of three different images. In particular formulas based on ratios (A/B), three components ratios (A/B/C or A*B/C) and ratio with additions such as 1/(A+B) or A/(B+C), or subtractions such as 1/(A-B) and A/(B-C) are most useful. Fluorescence images obtained at excitation/emission ranges: (250-300)/(300-400), 320/(350-450), (340-360)/(400-500), (400-450)/500-600 are sensitive to protein synthesis, protein content, extracellular proteolysis, cellular metabolism, as well as the structure and content of the extracellular matrix, including collagen and elastin contained therein. Reflectance images at 350-400 nm, 400-440 nm, 450-500 nm, 530-600 nm, and longer wavelengths as they are related to vascularity and proliferation, which is important in that vascularity changes in disease processes such as cancer and inflammation.

Formulaic video rate or still imaging provides maximal contrast between diseased tissue (lesions) and surrounding normal tissues, or between two different disease processes (such as cancer and inflammation), or between different grades of disease such as cancer, high grade dysplasia and low grade dysplasia (pre-cancer).

Synthetic formulaic images are computed with the goal to provide maximized contrast between lesions and surrounding normal tissue. Lesion refers to diseased tissue that includes but is not limited to cancer, pre-cancer, fibrosis, inflammation and ischemia. When implemented in endoscopy such computed formulaic images provides optimized lesion contrast, exceeding the contrast for normal white light visual observation without the need for labeling or manipulation of the tissue in a real-time manner.


  • Detection of precancerous lesions of the cervix and oral cavity
  • Visualization of adenocarcinoma in colon specimens


  • Not dependent on use of dyes
  • Formulaic images computed in real time
  • Easily integrated into endoscopes
  • Improves imaging of flat lesions that are easily missed
  • Well define lesion borders

Status: issued U.S. patent #9,952,157.
Further investigations will determine the efficacy of our methods in comparison to existing techniques. Investigators seek funding for instrument development as the current imager can only measure excised tissue as well as larger clinical study to obtain more data on small as well as flat lesions. This will extend their image library and help further optimize the image formula towards one that is most sensitive to small and flat lesions but also has high specificity.

Patent Information:
Contact For More Information:
Garrett Edmunds
Licensing Manager, UAHS-TLA
The University of Arizona
Lead Inventor(s):
Urs Utzinger
Bhaskar Banerjee
Timothy Renkoski
Nathaniel Rial
Logan Graves
image contrast optimization